PROJECT SUMMARY Recent improvements in the acute care of critically ill patients have led to new challenges in the management of patients who survive with persistent, severe organ failure after critical illness. The population of patients requiring long-term life support after critical care is rapidly growing, made up mostly of older adults, and associated with a large clinical and financial burden. One such intervention is the decision to place a gastrostomy tube to allow for provision of long-term artificial nutrition in patients who are unable to safely take in sufficient nutrition on their own due to new deficits in functional status, mental status, and respiratory function. However, basic knowledge of the epidemiology, drivers of gastrostomy use, and outcomes after gastrostomy placement is lacking. It is unknown, for example, how frequently patients are readmitted or seen in the emergency room after placement, or what the rate of mortality is one year after placement. The project proposed here seeks to quantify the incidence of gastrostomy tube placement across age groups in critically ill patients in the United States from 1993-2012, describe patient and hospital drivers of practice variation in the placement of gastrostomy tubes, and describe the determinants of patient-level outcomes after gastrostomy tube placement. We hypothesize that gastrostomy tube placement during critical illness has increased from 1993 through the late 2000s across all age groups, and that they now make up the predominant fraction of hospitalized patients receiving feeding tubes in older adults; we also hypothesize that utilization of gastrostomy tubes in the critically ill population vary at the hospital level. Finally, we hypothesize further that multiple determinants, including age, pre-critical illness frailty, and characteristics of acute critical illness can be used to predict rates of complications. The results will significantly advance our understanding of the current use of gastrostomy tubes among the critically ill as well identify the patient populations that may benefit the most from long-term artificial nutrition. Our overall goal is to provide the necessary foundation for a future shared- decision making tool that can be used by critically ill patients, their surrogate decision makers, and clinicians to help decide whether the likely outcomes of a gastrostomy tube are in line with the patient?s wishes and values. As part of the research training program, the principle investigator will complete a Masters of Science in Epidemiology in order to obtain the necessary epidemiologic, biostatistics, and health services research skills. This research project will be performed under the guidance of two health services investigators with expertise in managing large administrative databases as well as with advice from two experts in research on aging. !